Application of generative language models to orthopaedic practice.

OBJECTIVE: To explore whether large language models (LLMs) Generated Pre-trained Transformer (GPT)-3 and ChatGPT can write clinical letters and predict management plans for common orthopaedic scenarios. DESIGN: Fifteen scenarios were generated and ChatGPT and GPT-3 prompted to write clinical letters and separately generate management plans for identical scenarios with plans removed. MAIN OUTCOME MEASURES: Letters were assessed for readability using the Readable Tool. Accuracy of letters and management plans were assessed by three independent orthopaedic surgery clinicians. RESULTS: Both models generated complete letters for all scenarios after single prompting. Readability was compared using Flesch-Kincade Grade Level (ChatGPT: 8.77 (SD 0.918); GPT-3: 8.47 (SD 0.982)), Flesch Readability Ease (ChatGPT: 58.2 (SD 4.00); GPT-3: 59.3 (SD 6.98)), Simple Measure of Gobbledygook (SMOG) Index (ChatGPT: 11.6 (SD 0.755); GPT-3: 11.4 (SD 1.01)), and reach (ChatGPT: 81.2%; GPT-3: 80.3%). ChatGPT produced more accurate letters (8.7/10 (SD 0.60) vs 7.3/10 (SD 1.41), p=0.024) and management plans (7.9/10 (SD 0.63) vs 6.8/10 (SD 1.06), p<0.001) than GPT-3. However, both LLMs sometimes omitted key information or added additional guidance which was at worst inaccurate. CONCLUSIONS: This study shows that LLMs are effective for generation of clinical letters. With little prompting, they are readable and mostly accurate. However, they are not consistent, and include inappropriate omissions or insertions. Furthermore, management plans produced by LLMs are generic but often accurate. In the future, a healthcare specific language model trained on accurate and secure data could provide an excellent tool for increasing the efficiency of clinicians through summarisation of large volumes of data into a single clinical letter.

The cost effectiveness of potential risk factors for developmental dysplasia of the hip within a national screening programme.

Early detection of developmental dysplasia of the hip (DDH) is associated with improved outcomes of conservative treatment. Therefore, we aimed to evaluate a novel screening programme that included both the primary risk factors of breech presentation and family history, and the secondary risk factors of oligohydramnios and foot deformities. A five-year prospective registry study investigating every live birth in the study's catchment area (n = 27,731), all of whom underwent screening for risk factors and examination at the newborn and six- to eight-week neonatal examination and review. DDH was diagnosed using ultrasonography and the Graf classification system, defined as grade IIb or above or rapidly regressing IIa disease (≥4o at four weeks follow-up). Multivariate odds ratios were calculated to establish significant association, and risk differences were calculated to provide quantifiable risk increase with DDH, positive predictive value was used as a measure of predictive efficacy. The cost-effectiveness of using these risk factors to predict DDH was evaluated using NHS tariffs (January 2021). The prevalence of DDH that required treatment within our population was 5/1,000 live births. The rate of missed presentation of DDH was 0.43/1000 live births. Breech position, family history, oligohydramnios, and foot deformities demonstrated significant association with DDH (p < 0.0001). The presence of breech presentation increased the risk of DDH by 1.69% (95% confidence interval (CI) 0.93% to 2.45%), family history by 3.57% (95% CI 2.06% to 5.09%), foot deformities by 8.95% (95% CI 4.81% to 13.1%), and oligohydramnios nby 11.6% (95 % CI 3.0% to 19.0%). Primary risk factors family history and breech presentation demonstrated an estimated cost-per-case detection of £6,276 and £11,409, respectively. Oligohydramnios and foot deformities demonstrated a cost-per-case detected less than the cost of primary risk factors of £2,260 and £2,670, respectively. The inclusion of secondary risk factors within a national screening programme was clinically successful as they were more cost and resource-efficient predictors of DDH than primary risk factors, suggesting they should be considered in the national guidance.

Use of the behaviour change wheel to improve everyday person-centred conversations on physical activity across healthcare.

BACKGROUND: An implementation gap exists between the evidence supporting physical activity in the prevention and management of long-term medical conditions and clinical practice. Person-centred conversations, i.e. focussing on the values, preferences and aspirations of each individual, are required from healthcare professionals. However, many currently lack the capability, opportunity, and motivation to have these conversations. This study uses the Behaviour Change Wheel (BCW) to inform the development of practical and educational resources to help bridge this gap. METHODS: The BCW provides a theoretical approach to enable the systematic development of behaviour change interventions. Authors followed the described eight-step process, considered results from a scoping review, consulted clinical working groups, tested and developed ideas across clinical pathways, and agreed on solutions to each stage by consensus. RESULTS: The behavioural diagnosis identified healthcare professionals' initiation of person-centred conversations on physical activity at all appropriate opportunities in routine medical care as a suitable primary target for interventions. Six intervention functions and five policy categories met the APEASE criteria. We mapped 17 Behavioural Change Techniques onto BCW intervention functions to define intervention strategies. CONCLUSIONS: This study uses the BCW to outline a coherent approach for intervention development to improve healthcare professionals' frequency and quality of conversations on physical activity across clinical practice. Time-sensitive and role-specific resources might help healthcare professionals understand the focus of their intervention. Educational resources aimed at healthcare professionals and patients could have mutual benefit, should fit into existing care pathways and support professional development. A trusted information source with single-point access via the internet is likely to improve accessibility. Future evaluation of resources built and coded using this framework is required to establish the effectiveness of this approach and help improve understanding of what works to change conversations around physical activity in clinical practice.

What do healthcare professionals want from a resource to support person-centred conversations on physical activity? A mixed-methods, user-centric approach to developing educational resources.

Objectives: Healthcare is a fundamental action area in population efforts to address the global disease burden from physical inactivity. However, healthcare professionals lack the knowledge, skills and confidence to have regular conversations about physical activity. This study aimed to: (1) understand the requirements of healthcare professionals and patients from a resource to support routine physical activity conversations in clinical consultations and (2) develop such a resource. Methods: This study used codesign principles across two phases, actively involving relevant stakeholders in an iterative development process. The preparatory phase included a scoping literature review and workshops with multidisciplinary healthcare professionals and patients. The Delphi phase included the development of a draft resource, a three-stage modified online Delphi study and an external review. Results: The scoping review highlighted the importance of addressing time restrictions, a behaviour change skill deficit, the need for resources to fit into existing systems and meeting patient expectations. Consultation included 69 participants across two clinical workshops. They recommended using the internet, valued guidance on all aspects of physical activity conversations and were concerned about how to use a person-centred approach. The Delphi phase, including 15 expert participants, met agreement criteria in two stages to develop the resource. Conclusion: This mixed-methods study delivered an online resource that was codesigned with and based on the requirements of healthcare professionals and patients. The resource presents condition-specific '1-minute', '5-minute' and 'more minute' person-centred and evidence-based conversation templates on physical activity in an accessible and usable format to meet the needs of real-life clinical practice.

The articularis genu muscle and its relevance in oncological surgical margins.

AIMS: The aticularis genu (AG) is the least substantial and deepest muscle of the anterior compartment of the thigh and of uncertain significance. The aim of the study was to describe the anatomy of AG in cadaveric specimens, to characterize the relevance of AG in pathological distal femur specimens, and to correlate the anatomy and pathology with preoperative magnetic resonance imaging (MRI) of AG. METHODS: In 24 cadaveric specimens, AG was identified, photographed, measured, and dissected including neurovascular supply. In all, 35 resected distal femur specimens were examined. AG was photographed and measured and its utility as a surgical margin examined. Preoperative MRIs of these cases were retrospectively analyzed and assessed and its utility assessed as an anterior soft tissue margin in surgery. In all cadaveric specimens, AG was identified as a substantial structure, deep and separate to vastus itermedius (VI) and separated by a clear fascial plane with a discrete neurovascular supply. Mean length of AG was 16.1 cm ( ± 1.6 cm) origin anterior aspect distal third femur and insertion into suprapatellar bursa. In 32 of 35 pathological specimens, AG was identified (mean length 12.8 cm ( ± 0.6 cm)). Where AG was used as anterior cover in pathological specimens all surgical margins were clear of disease. Of these cases, preoperative MRI identified AG in 34 of 35 cases (mean length 8.8 cm ( ± 0.4 cm)). RESULTS: AG was best visualized with T1-weighted axial images providing sufficient cover in 25 cases confirmed by pathological findings.These results demonstrate AG as a discrete and substantial muscle of the anterior compartment of the thigh, deep to VI and useful in providing anterior soft tissue margin in distal femoral resection in bone tumours. CONCLUSION: Preoperative assessment of cover by AG may be useful in predicting cases where AG can be dissected, sparing the remaining quadriceps muscle, and therefore function.Cite this article: Bone Joint Open 2020;1-9:585-593.